Morbidity From Sleep Disturbance of Restless Legs Syndrome

RLS classification as a sleep-related disorder¹ emphasizes the sleep disturbance characterizing the disorder. By definition, rest provokes RLS symptoms and it can be seen as a disorder of that quiet drowsy state of waking (the predormitum) that precedes sleep but may occur at other times during the day. RLS will therefore generally disrupt sleep, and indeed, the population-based surveys indicate more than 75% of those with at least moderate RLS symptoms complain of significant sleep disturbance caused by their RLS.² It appears that some who have very mild RLS do not have any sleep complaint. In such cases, the patient appears to fall asleep fast enough to avoid developing significant symptoms while lying in bed; the longer periods of sitting in the evening, however, cause considerable distress. Most RLS patients, though, complain of interrupted sleep, insufficient hours of sleep, and difficulty falling and staying asleep. Curiously, only a minority complain of daytime sleepiness.² Figure 37-1 shows the range of RLS effects on sleep onset and sleep awakenings for those with moderate to severe RLS symptoms. These have been confirmed by polysomnographic studies of RLS patients³ and by reports from clinical trial populations showing decreased total sleep times, decreased sleep adequacy, and increased sleep disturbance, which are largely improved by treatment with the dopamine agonist ropinirole.^4,5 Sleep efficiency for the more severely affected RLS patient drops below 60% with less than 5 hours a sleep a night for most nights,⁶ producing what is probably the most extreme chronic sleep loss of any sleep-related disorder; yet, there have been few studies evaluating the effect of this sleep loss.

FIGURE 37-1. Restless legs syndrome (RLS)-related sleep complaints of those with moderate to severe RLS symptoms in a large primary care population survey. (A) Reported RLS-related sleep-onset times. (B) Reported RLS-related number of awakenings on usual nights.

From Hening W, Walters AS, Allen RP, et al. Impact, diagnosis and treatment of restless legs syndrome (RLS) in a primary care population: The REST (RLS Epidemiology, Symptoms, and Treatment) primary care study. Sleep Med 2004;5:237-246.

Profound daytime sleepiness should develop with this level of sleep loss, but surprisingly RLS patients generally do not report falling asleep in the daytime at inappropriate times (e.g., sitting in a car at a red light, talking on the telephone). Thus, the question arises—Aside from the discomfort from not sleeping, do RLS patients have any adverse effects from the sleep loss? That is, do RLS patients show changes in mood, cognition, attention, or pain perception generally associated with significant sleep loss? Unfortunately, to date, this has hardly been studied at all. Saletu and colleagues⁷ found no excessive sleepiness or vigilance decrement for untreated RLS patients compared with control subjects. Pearson and colleagues,⁸ however, did find that untreated RLS patients had decreased cognitive performance in the morning on verbal fluency and trail making tests consistent with the prefrontal cognitive effects shown for one night’s of total sleep deprivation.⁹ Aside from these studies, we know very little about the effects of the sleep disturbance in RLS. It looks as if it produces an awake, but possibly fatigued, individual with some disturbance of prefrontal cognitive functioning, a type of dysfunction also shown with attention deficit disorder.¹⁰

Quality of Life Measures Used in Restless Legs Syndrome

Three basic types of QoL measurements have been developed: general QoL assessments, disease-specific QoL measures, and judgments or recording of critical life-activities or events. The general QoL assessments usually focus on providing a broad-based assessment of health-related impacts on QoL. Several instruments have been developed to provide this type of QoL assessment, including the Nottingham Health Profile, the Sickness Impact Profile, and the 36-item Short-form Health Survey (SF-36). The most commonly used and best standardized is the Medical Outcomes Study¹¹ SF-36,^12,13 which has been shown to have very good psychometric properties when used in a patient population.¹⁴ The SF-36 also comes in two shortened forms: the SF-12¹⁵ and the SF-8.¹⁶ The SF-36 provides scores in eight dimensions and a summary score for four dimensions covering physical health (physical functioning, role physical, pain, and general health) and four covering mental health (role emotional, vitality, social functioning, and mental health).

The generic QoL allow quantitative comparisons with a normative population and between differing disorders. Thus, they provide a rationale for treatment and a possible metric for measuring its success. Unfortunately, the strength of the broad-based approach of these instruments also represents their major weakness. Many factors affect these scales so that when looking for treatment effects the change from treatment is obscured by changes or lack of other factors affecting QoL. Moreover, some of the QoL issues may be more specific to a particular disorder and as such need to be more directly addressed than occurs in a generic QoL. A disease-specific QoL can incorporate both dimensions of particular significance for a particular disorder and those that are expected to modulate with treatment of the disorder. Another targeted approach, the use of measures of life events or choices can focus on evaluating select and relevant behaviors or judgments about behaviors. This approach includes records such as work attendance or an inventory of patient judgments or choices about life events. They produce an interesting range of alternate QoL measures but have yet to become established as a standard approach.

RLS has now been evaluated using both the generic and disease-specific QoL. Both of these approaches focus on the adverse impact of health problems on daily functioning over the most recent 4 weeks. The generic QoL involve some generic health issues such as “bodily pain.” The RLS patient’s responses depend largely on how he or she relates the symptoms to these generic type descriptors. The disease-specific QoL instruments are not dependent on such a relation. Both measures focus on deficits and have a ceiling effect that could make it difficult to see enhancements to life that might occur with treatments, but the QoL model in this work focuses on evaluating the degree of disease related impairment and the improvements toward normal and not enrichment of daily life.

Quality of Life for Untreated Restless Legs Syndrome

The SF-36 has now been used in two population-based studies identifying individuals with RLS symptoms. The evaluation of RLS prevalence among elderly in Augsburg, Germany, used a standard clinical interview with a neurologic examination. The 369 participants completed the SF-36. RLS was diagnosed for 36 subjects (9.8%). Those with RLS compared with those without RLS showed on the SF-36 significantly (p <.05) lower scores for the mental health and depression dimensions. Although the vitality and role-physical scores were also lower for RLS, the differences were not statistically significant,¹⁷ but this was a small sample of older individuals (65 to 83 years old) who already have some compromise in physical health and vitality limiting somewhat sensitivity of the measure on these domains.

A larger population-based study used a validated set of questions to assess presence of RLS symptoms among 6014 subjects in the United States. The 187 subjects meeting criteria for having RLS symptoms also completed the SF-36 and rated their symptoms as either mild, moderate, or severe (including very severe).¹⁸ This population of those with RLS showed overall lower scores than those from an age- and gender-matched normal population on all domains except for role emotional¹⁹ (Fig. 37-2). Increasingly poorer scores on the SF-36 occurred for increasing severity of the symptoms (Fig. 37-3). Compared with the normative data, those with mild RLS symptoms showed significant differences only for energy/vitality; those with moderate RLS symptoms showed significant differences for all of the dimensions except role emotional; and those with severe RLS symptoms showed significant and dramatic impairments on all eight of the dimensions. There are some interesting differences between the American and somewhat younger sample of adults compared with the German elderly. Both studies showed significant decreased QoL for the mental health dimension and also reduced scores in the physical health dimensions, but for the American study the greater differences occurred for the physical health scores of bodily pain and role physical. Thus, we have a picture of mild RLS affecting most vitality, but moderate to severe RLS showing not only decreased vitality but also equally decreased role physical and bodily pain. This somewhat surprising picture of RLS indicates those with moderate to severe symptoms have marked impairment of QoL involving both vitality and physical limitations and pain. The issue of pain in RLS has to a large extent been ignored, although in the REST population-based study 59% reported pain with the symptoms and in another survey 56% of clinical patients with RLS similarly reported pain with RLS.²⁰ Thus, the sensory discomfort of RLS appears to be both intense and at least as significant as the loss of energy or vitality for effects on QoL (see Chapter 27). This issue is of course confounded by the fact that sleep loss lowers pain thresholds; the sleep effects of RLS may be contributing not only to decreased vitality but also to increased pain. However, sleep loss by itself seems an inadequate explanation for the marked deficits seen on these dimensions of the SF-36 in RLS.

FIGURE 37-2. SF-36 for all with restless legs syndrome symptoms compared with age- and gender-matched norms.

From Allen RP, Walters AS, Montplaisir J, et al. Restless legs syndrome prevalence and impact: REST General Population Study. Arch Intern Med 2005;165:1286-1292.

FIGURE 37-3. SF-36 domain scores for various levels of restless legs syndrome severity (mild, moderate, and severe, blue bars) compared with population norms (blue diamonds connected by lines).

Data from study reported by Allen RP, Walters AS, Montplaisir J, et al. Restless legs syndrome prevalence and impact: REST General Population Study. Arch Intern Med 2005;165:1286-1292.

In addition, the SF-36 scores of those with mild to severe RLS symptoms in the American population–based study show similar impairment to that of other significant chronic diseases (Figs. 37-4 and 37-5). Moderate to severe RLS affects SF-36–based QoL as much as depression, osteoarthritis, congestive heart failure, or type 2 diabetes.¹⁹

FIGURE 37-4. SF-36 domain scores for restless legs syndrome (RSL) compared with cardiovascular problems. The RLS scores are as low or lower than those for major cardiac conditions.

Rights were not granted to include this figure in electronic media. Please refer to the printed book.

From Kushida C, Martin M, Nikam P, et al. Burden of restless legs syndrome on health-related quality of life. Qual Life Res 2007;16:617-624.

FIGURE 37-5. SF-36 domain scores for restless legs syndrome (RSL) (in blue) compared with other chronic major medical conditions. The RLS scores are as low or lower on most domains compared with depression, osteoarthritis, type 2 diabetes, or chronic obstructive pulmonary disease.

Rights were not granted to include this figure in electronic media. Please refer to the printed book.

From Kushida C, Martin M, Nikam P, et al. Burden of restless legs syndrome on health-related quality of life. Qual Life Res 2007;16:617-624.

These data provide strong support for the need to provide treatment for RLS.

Quality of Life for Treated Restless Legs Syndrome

Treatment effects on QoL are usually more clearly seen for disease-specific QoL measurements. Three such QoL scales have been developed. The first of these, developed at Johns Hopkins University and referred to as either the Hopkins RLSQoL or simply the RLSQoL, used expert opinion and patient support groups to develop a 18-item questionnaire covering a wide range of reported adverse effects of RLS on daily living. This test was found to have a valid total score with good psychometric properties (Chronbach alpha = 0.92, test-retest =0.84). The total score showed significant differences with even small changes in subjective report of symptoms occurring over a few weeks suggesting adequate responsiveness for treatment evaluation.²¹ This questionnaire was further evaluated in a two large clinical trial populations where the total score was again found to have very good psychometric properties and reasonable correlations of 0.67 or better with independent RLS severity ratings.²²

The Johns Hopkins RLSQoL was used in two large placebo-controlled clinical trials evaluating benefits of a dopamine agonist (ropinirole) after 12 weeks of treatment. In both trials ropinirole showed an improvement in the RLSQoL score 25% greater than placebo.^4,23 The degree of improvement was reasonably impressive, exceeding a 1 standard deviation change, which would be considered a large and clinically meaningful improvement in QoL. This has been the most widely used RLS specific quality of life scale. It has been used in most clinical trials and is available in well-validated translations in many languages from http://www.mapi-research.fr/.

The generic QoL SF-36 was also evaluated in these two large clinical trials and as expected showed less treatment effects than did the disease specific Johns Hopkins RLSQoL. In one of these trials conducted in 10 countries in Europe there were no significant changes noted in the SF-36 scores with treatment.²³ The other clinical trial from centers in Europe, Australia, and North America, however, showed significantly greater improvement for ropinirole compared with placebo on 3 of the SF-36 domains, i.e., mental-health, social functioning, and vitality. The first two of these domains were not the ones with the biggest difference between those with and those without RLS symptoms, so why did these show responsiveness when the domains of role-physical and pain did not? One explanation may be furnished by study design: this study selected patients with only evening and nighttime symptoms and gave medication in one daily dose before bedtime with the goal of reducing late night symptoms and improving sleep. Although these were selected to be moderate to severe patients, the limitation of no daytime symptoms probably excluded the more severe RLS and thus the domains of pain and role-physical on the SF-36 were less likely to be reduced. They were, therefore, not the domains most expected to show treatment benefits. In contrast the study demonstrated the expected significant treatment benefit on sleep with improvement in the vitality domain. These differences regarding the significance of the pain and role-physical dimensions of QoL for different RLS populations needs to be further explored. It may represent a severity effect or possibly reflect differing phenotypes.

The second disease-specific QoL scale was developed based on a large convenience sample from the Internet of RLS patients, most of whom were receiving treatment. The diagnosis of these patients relied on their self-report, which was unlike the Johns Hopkins population, where the diagnosis was made by an experienced clinician. This second scale, referred to as the RLSQoL-1, has four distinguishable factors that were labeled: daily function, social function, sleep quality, and emotional well-being.²⁴ This 17-item scale also has very good psychometric properties and it was used in one study of RLS patients, many of whom continued to have symptoms even if treated. The RLSQoL-1 results when analyzed in detail in relation to subjective report of sleep status indicated that a large part of the RLS impact on functioning was associated with the degree of sleep disturbance. Once the sleep status was considered, there appeared to be no further relation between the RLS severity and impact on functioning.²⁵ This was a limited study, but it raised the possibility that the sleep disturbance produces most of the RLS impact on QoL. Again, however, the major dimensions of pain, discomfort, and role-physical were not adequately covered in any of the material used in this study.

The third specific QOL scale is the Restless Legs Syndrome Quality of Life (QoL-RLS) questionnaire.²⁶ From a survey on 721 patients of North-German RLS advocacy groups, the item pool was determined by a content analysis of written reports from RLS patients on the consequences of their RLS disorder on everyday life functioning. The items are to be answered on a 6-point scale (0 = no impairment to 6 = very severe impairment). Factor analysis of a questionnaire combining the items resulted in one scale with 12 items. The most relevant item consists of a global rating of quality of life “all in all.” Further items are related to consequences of the RLS symptoms on sleep, activities of daily living, mood, and social interactions, to the consequences of disturbed sleep on everyday life and tiredness during the day and on mood as well as to consequences of pain and side effects of treatments on daily activities. A final section of the questionnaire asks for evaluation of coping behavior (effort of measures to get relief from the symptoms, avoidance of social situations, changes in lifestyle). Reliability estimates showed favorable psychometric qualities of the QoL-RLS (Cronbach’s α=.954). Intercorrelations of the total score from this QoL-RLS with RLS-6 severity scales (at time falling asleep, at night, during the day at rest, and during the day when active) were sufficiently high (around r = 0.60); however, the correlation to the total score of the International Restless Legs Syndrome Study Group (IRLSSG) Rating Scale was exceptionally high (r = 0.925), indicating a presumably high affinity of the IRLS scale to quality of life. The QoL-RLS was applied in several clinical trials; in all these trials, the total score was more improved by active dopamine agonist therapy than by placebo^27,28 or by levodopa.²⁹ The instrument is available in several languages.

Conclusion

The morbidity of the RLS sleep disturbance appears to include cognitive deficits similar to those seen with one night’s sleep loss but does not include significant daytime sleepiness, suggesting there may be a hyperarousal with RLS during the daytime. Overall, the impact of the RLS sleep disturbance has unfortunately not been adequately studied. One limited study, however, suggested it was a primary contributor to the RLS effects on QoL.

QoL studies of RLS have proved more useful for both of the two expected areas of determining the need and indications for treatment. First, they demonstrated that RLS significantly impacts QoL as much or more than other chronic diseases. The results strongly support treatment for those with moderate to severe symptoms with the need for treatment of mild symptoms being less certain. The studies also demonstrate that both the sleep-related dimensions such as energy/vitality and the sensory dimensions of role-physical and pain should be considered when evaluating and treating RLS patients. It may be that the RLS symptoms are more pervasive than suggested by the focus of most therapies on the nighttime. The pain and role physical aspects of RLS deserve to be explored more carefully.

The treatment results similarly show improved QoL on the disease specific QoL. The SF-36 results suggest some possible benefits for the sleep-related dimension of vitality and also for the mental health and social functioning but not for the role-physical or pain dimensions of the SF-36. This may reflect subject selection or again a failure to target treatment to address all aspects of RLS. Clearly, QoL evaluations should continue in future treatment trials. The SF-36 is widely available in multiple languages. The Johns Hopkins RLS-QoL has been translated into several languages and is available from MAPI-Research (http://www.mapi-trust.org/t_03_serv_dist_user.htm, where it is listed under J as the Johns Hopkins Restless Legs Syndrome Quality of Life questionnaire [RLS-QoL]). The QoL-RLS developed in Germany is available at http://www.kohnen@imerem.de.